Table 3

Score Distributions and Internal Consistency for Continuity Dimensions

Subscale (Items, Response Type, Range)	Item-Total Correlation Range	Cronbach α	Factor Loadings Across Response Format	Mean	SD	Median	Skew
Main clinician
Coordinator role (5 items, evaluative,^a 1 to 5)	0.65–0.74	0.87	0.60–0.66	3.65	0.88	3.76	−0.28
Comprehensive knowledge of patient (4 items, evaluative, 1 to 5)	0.70–0.78	0.89	0.74–0.79	3.70	0.86	4.0	0.49
Confidence and partnership (3 items, evaluative, 1 to 5)	0.71–0.76	0.86	0.76–0.79	3.9	0.79	4.0	−0.87
Various clinicians
Confidence in team (2 items, evaluative, 1 to 5)	0.67	0.80	0.49–0.57	3.04	1.12	6	−0.15
Role clarity and coordination within clinic (3 items, reporting,^b 0 to 3)	0.44–0.68	0.66	0.59–0.71^c	0.43	0.82	1	1.96
Role clarity and coordination between clinics (3 items, reporting, 0 to 3)	0.60–0.72	0.82	0.76–0.80	0.44	0.91	0	1.95
Information gap between clinicians (6 items, reporting, 0 to 6)	0.50–0.70	0.85	0.58–0.75	1.9	1.9	1	0.65
Patient as partner
Care plan (7 items, reporting, 0 to 7)	0.52–0.63 (dropped item 0.38)	0.81	0.41–0.65 (dropped item 0.31)	4.0	2.24	4	−0.26
Self-management information provided (4 items, evaluative, 1 to 5)	0.74–0.81	0.93	0.77–0.91^c	3.86	1.16	4	−0.96

↵a Evaluative subscales were scored by averaging the value of individual items; range reflects response scale. Coordinator role is a weighted average giving higher weight to more difficult items.
↵b Items in reporting subscales were first dichotomized to reflect presence of any problem (more than sometimes), then summed to reflect number of problems encountered.
↵c Estimates across all response formats are not reliable due to small sample size.